mirrors/gnss-sdr
1
0
Fork 0
mirror of https://github.com/gnss-sdr/gnss-sdr synced 2024-09-27 22:58:24 +00:00
Code Issues Releases Wiki Activity

add neon implementation

Browse Source
This commit is contained in:
Carles Fernandez 2016-01-23 21:22:30 +01:00
parent c5252da7fd
commit cd2f0b86f6
1 changed files with 48 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
src/algorithms/libs/volk_gnsssdr_module/volk_gnsssdr/kernels/volk_gnsssdr/volk_gnsssdr_16ic_x2_multiply_16ic.h
View File

@ -161,4 +161,52 @@ static inline void volk_gnsssdr_16ic_x2_multiply_16ic_u_sse2(lv_16sc_t* out, con
}
#endif /* LV_HAVE_SSE2 */
#ifdef LV_HAVE_NEON
#include <arm_neon.h>
static inline void volk_gnsssdr_16ic_x2_multiply_16ic_neon(lv_16sc_t* out, const lv_16sc_t* in_a, const lv_16sc_t* in_b, unsigned int num_points)
{
lv_16sc_t *a_ptr = (lv_16sc_t*) in_a;
lv_16sc_t *b_ptr = (lv_16sc_t*) in_b;
unsigned int quarter_points = num_points / 4;
int16x4x2_t a_val, b_val, c_val;
int16x4x2_t tmp_real, tmp_imag;
unsigned int number = 0;
for(number = 0; number < quarter_points; ++number)
{
a_val = vld2_s16((int16_t*)a_ptr); // a0r|a1r|a2r|a3r || a0i|a1i|a2i|a3i
b_val = vld2_s16((int16_t*)b_ptr); // b0r|b1r|b2r|b3r || b0i|b1i|b2i|b3i
__builtin_prefetch(a_ptr + 4);
__builtin_prefetch(b_ptr + 4);
// multiply the real*real and imag*imag to get real result
// a0r*b0r|a1r*b1r|a2r*b2r|a3r*b3r
tmp_real.val[0] = vmul_s16(a_val.val[0], b_val.val[0]);
// a0i*b0i|a1i*b1i|a2i*b2i|a3i*b3i
tmp_real.val[1] = vmul_s16(a_val.val[1], b_val.val[1]);
// Multiply cross terms to get the imaginary result
// a0r*b0i|a1r*b1i|a2r*b2i|a3r*b3i
tmp_imag.val[0] = vmul_s16(a_val.val[0], b_val.val[1]);
// a0i*b0r|a1i*b1r|a2i*b2r|a3i*b3r
tmp_imag.val[1] = vmul_s16(a_val.val[1], b_val.val[0]);
// store the results
c_val.val[0] = vsub_s16(tmp_real.val[0], tmp_real.val[1]);
c_val.val[1] = vadd_s16(tmp_imag.val[0], tmp_imag.val[1]);
vst2_s16((int16_t*)out, c_val);
a_ptr += 4;
b_ptr += 4;
out += 4;
}
for(number = quarter_points * 4; number < num_points; number++)
{
*out++ = (*a_ptr++) * (*b_ptr++);
}
}
#endif /* LV_HAVE_NEON */
#endif /*INCLUDED_volk_gnsssdr_16ic_x2_multiply_16ic_H*/